Research On Resource Scheduling Mechanism Based On OpenStack

Cloud computing technology in the era of large data completely subvert the traditional data computing storage model,and it is a new type of computing service model.And more and more individuals become the users of the cloud computing,the size of the data center is also growing,the problem of the low utilization of resource and high energy consumption is serious.Then how can we use resources efficiently,reduce energy consumption and improve service quality for the large data center is essential,it is an important method to improve the performance of the entire cloud computing server cluster by the improvement of the cloud computing platform.In this paper,the mainstream open source cloud computing platform such as OpenStack is studied in depth,and the research and improvement of its existing resource scheduling mechanism are carried out.The main work is composed of two parts: the improvement of virtual machine initial placement mechanism and virtual machine dynamic migration mechanism.1.A virtual machine initial placement mechanism based on resource performance perception is proposed.First of all,the pseudo and excess use of physical resources,the excess use of multiple set at 1.2 times,and then put forward the matching degree concept of the resource performance of physical machines and virtual machines.In the matching weight calculation,we taking into account the CPU,memory,storage and bandwidth,and according to the service type,we consider the different weight coefficient in the formula of calculating the weight value.Secondly,we put forward the concept of expectation of resource withdrawal in the future of all virtual machines on the physical machine,and finally we summed up the three factors of the matching degree of resource performance and the expectations of retreatment of physical resource and the resource utilization rate of physical machine to calculate the probability of virtual machine placed on a physical machine.We select the most appropriate physical machine to place the virtual machine,and through the experimental verification based on CloudSim,that is a cloud computing platform: virtual machine initial placement algorithm MHD-VMP performed well in the three areas of the number of running hosts,the average resource utilization of CPU of the running host and energy consumption,in contrast to OpenStack-Initial algorithm,the improvement of proportion of resource utilization is up to 10.46% in the optimal case,so it is suitable for the initial placement of virtual machines in the OpenStack cloud platform generally.2.A dynamic migration mechanism of virtual machine based on load perception is proposed.This mechanism senses the dynamic load changes of the cloud data center,balancing the data center load through the dynamic migration of the virtual machine.In the process of dynamic migration of virtual machines,this paper consider three aspects: when to migrate virtual machine,which virtual machine should be selected to migrate and where the virtual machine can be located,that is the target physical machine.In this paper,the triggering migration algorithm of virtual machine which named LRR-TMM,the selection algorithm of virtual machine which named MMC-SM and the selection algorithm of target physical machine which named PR-DPS are proposed for the 3w process,and then we take into account the three-part to propose the dynamic migration algorithm of virtual machine which named LA-VMDM based on load perception,it can improve the effectiveness of energy and ensure the stability and sustainable service of the data center while ensuring the provision of quality service.